Ned,
I have no problem with your original assertion that evolution doesn't predict increasing complexity. I was only trying to say that the amusing thing, to me, on the web site I mentioned was the idea that Eohippus had to be more complex than a modern horse because it had more toes.
I found Lyn Margulis' point that mitochondria are really the most successful life form rather interesting. While I haven't read enough to understand all her ideas on the subject, I do find it hard to argue that if mitochondria were once a separate life form, they sure found a great survival mechanism through making almost all later life dependent on them, even most bacteria.
We're off topic, but since I'm the only one who gave you another web site quote, I guess all the other posts were off topic, too.
But, since I'm here, how about this one. John Woodmorappe made a list of 350 discrepant radiometric dates from published literature. Glenn Morton then answered Woodmorappe by posting a graph of Woodmorappe's dates, which show that the dates, all chosen by Woodmorappe because they are more than 20% off, still go in a generally correct linear direction by age.
This is the hilarious part. Woodmorappe, in order to defang Morton, writes an entire reply attacking the data! It's Woodmorappe's own data, but one of his lines is, "Finally, it is the very selective publication of obtained isotopic dates (a fact that Morton acknowledges) that makes any would-be comparison of the numbers of good versus bad dates totally meaningless, nullifying his cheap shot argument all the more."
Well, it may nullify Morton's response, but it most certainly nullifies Woodmorappe's original article. Why did he publish those dates, forcing Morton's reply, if his compilation of bad dates is totally meaningless (which was Morton's point, anyway).
Sigh...Oh, Woodmorappe's article is at
http://www.trueorigin.org/ca_jw_02.asp.